Drift in the history of English

There is a generalisation about the development of English morphology which holds true particularly for the Middle English period: inflections are reduced and then lost with other mechanisms arising which compensate for this. This gradual shift from a synthetic language type in Old English to a analytic type in Modern English has been known as drift ever since the term was introduced by the American linguist Edward Sapir at the beginning of this century.

Three reasons for drift in English

1) Simplification, lack of redundancy in bilingual situations (Old English and Scandinavian in the north of the country)

2) Infection of Old English through contact with British Celtic which already had inflectional decay due to phonetic attrition.

3) Long-term effect of the initial stress-accent of Germanic which led to a corresponding weakening of unstressed syllables, typically morphological endings.

Not all languages which were once synthetic have lost their inflections. For instance verbal and nominal endings have been retained to a large extent in German. Hence the question arises as to whether there are recognisable causes for the observed drift in English. The standard wisdom on the subject is that the mixture of languages in the Old English period gave the impetus. Certainly there was a high degree of bilingualism in the north as of the 9th century when the Scandinavians established a firm foothold there. Furthermore, it is true that the north of England has been more innovative in morphology than the south, in keeping with the assumption that mixing leads to change, given that when speakers of one language are dealing with those of another then they tend to leave out unnecessary, redundant elements which are not required for conveying meanings.

There may well have been a further powerful force behind the demise of inflections in Old English. It is known that the Celtic languages which were in England before the arrival of the continental Germanic tribes showed considerable weakening in the articulation of consonants — a phonetic process known as lenition — and that this led to the ultimate loss of many inflections in these languages (Modern Welsh and Cornish and Modern Irish in Ireland). Now the Germanic tribes mixed with the Celts once they had subdued them. They did not banish them out of the land they conquered, as shown by the Old English word wealh which means both ‘Celt’ and ‘slave’. The assumption here is that the ‘soft’ pronunciation of consonants in British Celtic infected the pronunciation of Old English — in the north as much as elsewhere in the country as Celtic was spoken over the entire country and not just in the west and south-west as was later to be the case with the retreat to these areas. This is not an unreasonable assumption as it is known from other regions that linguistic features — particularly low-level ones, such as lack of aspiration with consonants or the use of extra vowels to break up groups of consonants — tend to cluster in geographical areas. The advantage of the hypothesis of an influence of Celtic on Old English is that it gives a greater time depth to the decay of inflections. The fact that this influence is not immediately apparent in Old English has to do with the fact that the West Saxon koiné masked many of the innovations which had taken place in morphology and which only came to the surface in writing in the documents of the Early Middle English period — as of the late 12th century when written material in English rather than French or Latin began slowly to appear again.

There is a third reason which should be taken into account here. If one compares the Germanic sub-group of languages with others in the Indo-European family then one sees that it developed a strong initial stress-accent in its early development, probably sometime after the first millennium BC. It shares this stress pattern with Italic but not with the Slavic or Baltic subgroups for example. Stress frequently has a demarcative function in language and in Germanic it was placed on the lexical root of words so that these were easily recognisable. This must have been different before in Indo-European because the oldest stages of the language family have variable stress and within Germanic there are reflexes of the former stress pattern, for instance in the alternation between /s/ and /r/ in Verlust but verlieren which goes back to Verner’s Law combined with rhotacism (the shift of /z/ to /r/ in intervocalic position). The result of strong lexical root stress is the backgrounding of morphological affixes — prefixes and inflectional endings. The less distinct the phonetic profile of an ending is, the more likely it is to be neglected by speakers and eventually dropped or for it to merge with other endings. Thus neither German nor English still have a distinction between inflectional /-n/ and /-m/ although the former has retained a far greater number of inherited inflections.

As is often the case in language development the reasons for the drift towards an analytic type are probably manifold, then one would argue not so much about a single cause as the relative importance of the different causes. Whatever the weighting of the contributory factors in the drift, the effect was the same: phonetic attrition led to the blurring of inflectional distinctions and later to their loss. For instance the Old English nominal endings -a, -u, -e originally had separate vowel values but these coalesced to one in Middle English, -e [ə]. The distinctions between /-n/ and /-m/ in endings was lost and this along with the vowel mergers meant that -an, -on, -un, -um all reduced to -en [ən]. A general levelling of the phonetic distinction between unstressed /a/ and /e/ meant that /-as/ and /-es/ as well as /-aθ/ and /-eθ/ were no longer distinguished in pronunciation. Final -e [ə] was lost in the north by the mid 13th century and by the end of that century had disappeared in the south as well.